Hello, I noticed that the training of TransE is very slow. Upon closer inspection, I found that the default GPU ID in the code is set to 1. This affects researchers who do not have a second GPU, causing the training to automatically fall back to CPU. I hope you can correct this promptly. Thank you.
